# 🌍 Terraforming-Planet image-gen
### An image generation platform for visual learning and experimentation in terrain formation, self-sustaining photovoltaic vehicles, and planetary engineering
# “Visualize first engineer wisely ,every planet is a system. Every system can be understood.”
**Solution technology for the Planet**
Terraforming-Planet is an educational project focused on terrain formation, water retention, and photovoltaic vehicles.
We combine science with visualization to support land restoration, environmental protection, and energy transition.
**Try the test image generation at this link**
# 👉 **[Click here and test the image generation model](https://terraformingplanet.terraforming-planet.workers.dev/)**
Below is a screenshot from image generator tests:
---
“AI is not the answer. It’s the tool we use to ask better questions.”
“Technology serves best when it teaches before it transforms.”
“Learning planetary engineering through visualization.”
# 🌍 Terraforming Planet — Image Generation Lab
> **Experimental AI image generation laboratory**
> An educational and engineering project demonstrating how **GPT Image + Codex + Cloudflare Workers**
> can be used to learn terraformating, terrain shaping, and future-oriented technology design.
---
## ✨ Project Idea
**Terraforming Planet** is a hands-on experiment showing the complete process of building
an AI image generator — from concept, through code architecture, to a working web application.
The project does not focus solely on image aesthetics.
Its goal is to **understand processes**:
- planetary terraforming,
- terrain shaping (mountains, valleys, deserts, oceans),
- energy utilization (e.g. photovoltaic machines),
- and using AI as an engineering and educational tool.
---
## 🧠 How was the generator built?
1. **Concept**
Terraforming visualization as a learning and analysis tool.
2. **Codex**
Iterative design of repository structure and code refinement.
3. **GPT Image API**
Generation of realistic images (returned as base64).
4. **Cloudflare Workers**
Secure API — the OpenAI key never reaches the frontend.
5. **Vite + Vanilla JS**
Lightweight, fast frontend without heavy frameworks.
6. **GitHub Actions**
Automated build and deployment.
The repository documents the **entire creation journey**, not just the final result.
---
## 🧱 Repository Structure
/apps
├─ web/ → Frontend (Vite + Vanilla JS)
│ ├─ src/
│ │ ├─ styles/ → main UI styles
│ │ └─ assets/ → SVGs, icons, backgrounds
│ └─ dist/ → production build
│
└─ worker/ → Image generation API
└─ Cloudflare Workers + OpenAI
---
🎨 What does the app offer?
🧪 Image laboratory
✍️ Prompt editor (style, format, size)
🖼️ AI image generation
📋 Prompt copying and result analysis
⚡ Rendering via data_url (no temporary links)
---
🚀 Local Setup
Frontend
cd apps/web
npm install
npm run dev
Address:
http://localhost:5173
---
Worker (Cloudflare)
cd apps/worker
npm install
npx wrangler deploy
In Cloudflare Dashboard → Worker → Settings → Variables
Add Secret:
OPENAI_API_KEY
Tests:
GET /health
POST /generate
Body:
{
"prompt": "a solar excavator terraforming desert",
"size": "1024x1024"
}
> GPT Image models return base64 — the worker maps it to data_url.
---
🌐 Web ↔ Worker Configuration
.env file in apps/web:
VITE_API_BASE=https://your-worker.workers.dev
VITE_WORKER_URL=https://your-worker.workers.dev
---
📦 Deployment (GitHub Pages)
Workflow:
.github/workflows/pages.yml
Automatically:
builds apps/web,
publishes apps/web/dist,
sets environment variables.
Requirements:
1. Correct base in vite.config.js
2. GitHub Pages → Source: GitHub Actions
3. Push to main
---
☁️ Deployment (Cloudflare Pages – optional)
cd apps/web
npm install
npm run build
Settings:
Build command: npm run build
Output: apps/web/dist
Root directory: apps/web
---
🔑 Environment Variables
Variable Description
OPENAI_API_KEY API key (worker only)
VITE_API_BASE Worker API address
VITE_WORKER_URL Worker URL alias
---
❓ FAQ
I don’t see an image after clicking “Generate”.
Check the worker and VITE_API_BASE.
“Missing OPENAI_API_KEY” error.
Check the Secret in Cloudflare Worker.
The page does not start.
Run npm install in apps/web.
---
🌱 Why this project?
AI as an engineering tool, not magic
Learning through visualization
Real architecture: frontend + worker + API
Ready to fork and extend
---
🤝 Community & OpenAI
Created as part of the Community Dev Challenge
and open for further experimentation.
👉 Online generator:
🔗 https://terraformingplanet.terraforming-planet.workers.dev/
---
Made with ☀️ AI, ⚙️ engineering, and 🌍 future-focused thinking
---
This is ONE FILE.
If anything changes — we edit THIS, not add new ones.
---
The project serves education and the common good.
We encourage collaboration and knowledge sharing.
Image generation model
Terraforming Planet • Cloudflare Worker
Image Generator (OpenAI)
example screen
https://github.com/user-attachments/assets/2056e93e-3a26-46dd-bf37-fd376d03bc29
https://github.com/user-attachments/assets/ac951fc0-bcb3-4261-8d82-3ef418f17b65
We create open demos and tools
that turn terraforming ideas into images, concepts, and educational processes.
From valleys and mountain ranges to megastructures — explored through image generators and well-designed prompts.
🌐 Home • 🧪 Demos • 🎯 Mission • 🧠 How generators help learning • 🤝 Collaboration • ⚖️ Responsibility
🚀 What is this project?
Terraforming-Planet is an organization focused on learning, prototyping, and storytelling around:
terrain formation (mountains, valleys, deltas, basins),
futuristic construction and photovoltaic machines,
planetary-scale engineering — presented visually.
Main idea:
image generators + good prompts = fast exploration of engineering concepts
without building heavy simulations or 3D pipelines.
🎯 Mission
We aim to:
Teach how terrain is shaped (geology + engineering thinking),
Prototype landscape-forming machines (e.g. PV excavators, autonomous builders),
Visualize “what-if” scenarios for planets and ecosystems,
Connect people: artists, developers, and engineers working for the common good.
Why it matters: images simplify complex systems and make them easier to understand, analyze, and develop.
🧠 How image generators help learning terraforming
Generative graphics can act as an educational laboratory:
1) Rapid hypothesis testing
Instant variations:
stronger / weaker gravity,
different planetary crust materials,
erosion intensity,
water cycles and river deltas,
construction strategies (terraces, canals, dams).
2) Concept iteration and communication
A single “4-in-1” image can show:
initial state → intervention → intermediate stage → final result
perfect for documentation, learning, and discussion.
3) Designing machines together with the environment
Machines evolve alongside the landscape they create:
PV excavators carving valleys,
autonomous vehicles stabilizing terrain,
modular systems building mountains and flood barriers.
4) Prompt-based scientific thinking
Good prompts enforce:
constraints,
measurable effects,
process stages,
coherent framing and perspective.
This project is open to discussion, ideas, and constructive feedback.
👉 Leave a comment or start a discussion here:
🔗
👉 Add a comment or share generated images
You can:
- ask questions
- share ideas
- report issues
- propose improvements
- discuss scientific or engineering concepts
Every voice helps improve the project 🌍
Example graphic drawn in GIMP used to train an educational OpenAI AI model (copy)
xxx.
